Last year I visited the BMW-MINI Production Plant, at Cowley, Oxford.
This was arranged as part of our theme on 'innovation' for Warwick
Business School's Innovation network. The bodyshell assembly hall is
90% automated. Hundreds of huge robots, clustered in protective
'cells' working completely autonomously, welding, pressing and
riveting panels. The robots maintain themselves, with 'slaves'
sharpening arc welders, switching tool assemblies and calling for
parts refills. The few humans we spotted were driving fork-lift trucks
and feeding the robots' insatiable appetite for materials.
Manufacturing has been at the forefront of using AI and robots to
improve productivity. Robots don't get tired, they don't need
meal-breaks or holidays and can work 24 x 7 if required. They also
offer more accuracy and precision than human workers, but as yet, not
agility. However, recent advances are freeing robots from their cages
and enabling humans to work alongside them. For example, the robot
providing the strength and precision to insert the dashboard console
into the car, then moving away while the human does any fiddly bits
requiring agility. Welcome to the world of 'cobots'.
Yes, manufacturing has come a long way in the past 60 or so years
from where the first picture (below was) taken.
Can we say the same about office work? The left-hand picture below is
a typical 1950's office environment. What has changed when compared to
the current office environment? Yes, we have computers on our desks
now, but not much else has changed in almost 70 years.
Ah, I hear you say, what about email? Or desktop applications, such
MSOffice and CRM? Harbingers of the paperless office! Have you seen
one? No, nor me. Do these really represent the best office innovations
of the past 70 years? Unfortunately it would appear so. The process
for updating a credit & debit spreadsheet is the same now as it
was 70 years ago, when accountants were manually filling in ledgers.
Routine, repetitive and robotic.
There is much talk about 'digital
transformation', an IT term which means nothing to most office
workers, and even less in terms of how this will change or optimise a
business process. More recently, we hear about agile and
flexible working, and the
confusion about the two. At least these do offer some benefits
in terms of process optimisation and the workplace environment
(including working from home), but they do little in terms of changing
the actual work.
So, what's the answer? I think we've already seen it in the way that
manufacturing has grasped and utilised AI and robotics over past 10-15
years. This doesn't mean having a physical robot in the office, but
rather looking at business processes in terms of what a machine is
good at (speed, repetition, accuracy) and what a human is good at
(decision making, handling exceptions, social interactions,
collaboration, knowledge sharing) and intelligently
combining the two.
In other words, don't make a human do robotic tasks, and don't
expect a robot to act like a human!
I don't want to invent a lexicon, but I call this 'Intelligent
Automation', and is combination of Robotic
Process Automation (RPA) and Artificial
Intelligence (AI). I note vendors in this space refer to this as
creating a 'digital workforce', which (IMHO) is as meaningless as
RPA is a software robot that mimics human actions; AI is the
simulation of human intelligence by machines. Put another way, RPA is
associated with doing whereas AI and machine learning (ML) are
concerned with thinking and learning, respectively. According to Grand
View Research, the RPA market is projected
to be worth over $3 billion by 2025, and industries like health care,
telecommunications and manufacturing have all invested in RPA technology.
Typically, processes that are ready for RPA/AI can be defined as:
- Having high volume transactions or keyboard activities
- Part of a defined workflow
- Being error-prone due to manual entry
- Being speed-sensitive with the possibility of causing delay to
- Needing more than one system (e.g. requiring dual data entry)
- Requiring actions such as searching, collating, updating, matching
- Abstracting information from unstructured data (e.g. emails)
- Needing irregular labour resources (e.g. low and peak periods)
But what about the human? The office (knowledge) worker remains a
fundamental part of the process, but is no longer doing repetitive or
robotic tasks. Not everything in life is predictable, and the same
goes for business processes. There will be issues and exceptions which
require problem solving and interactions with customers or other
employees. The difference now is humans will have more time to do
these things, and to adapt and improve their skills and knowledge.
Intelligent automation is the enabler for the new augmented workforce.
The office revolution is coming - it's a shame that it's taken
almost 70 years to get here!